Marcelo Thompson, Assistant Professor at the University of Hong Kong, Faculty of Law, proposes an efforts-based approach to Internet intermediary liability. This blogpost is one of a series reflecting discussions held at the LSE on 8 July 2016 as part of a Media Policy Project workshop on “Digital Dominance: Implications and Risks”, a summary of which will be available on this website shortly.
In debates concerning Internet intermediary liability, an often-expressed view is that intermediaries (such as Facebook and Google) shouldn’t be turned into adjudicators, who reason and decide about the legal or illegal nature of content they host, and thus about whether or not to take such content down. But is that a plausible view?
Intermediaries, after all, necessarily must and will make such decisions in one way or another. Once notified of the existence of content that violates people’s privacy, reputation, or children’s rights, can intermediaries avoid weighing those rights against freedom of expression and vice-versa?
Sure, we could compel intermediaries to defer everything to the courts. Yet, courts don’t work in Internet time, geography, or economy. The consequence would be that, with content remaining online, freedom of expression would always win, and other rights lose.
But there is a second reason why we wouldn’t want to defer everything to courts. Isn’t it at the very core of any activity to make decisions that are central to it? And what is more central to being a host than making decisions about … hosting? Remove the reason element from any practice and we are left with a rather impoverished expression of it.
The real problem with Internet intermediaries isn’t having private actors making legal decisions. We all make decisions about right and wrong all the time, and the law is, ultimately, a living expression of the multitude of these decisions. The real problem with Internet intermediaries is rather how, with what diligence, public spirit, and, indeed, responsibility they make the decisions they make.
Yet, a concern with this ‘how’ is nowhere to be found in existing liability regimes. Rather, these regimes rely on arbitrary, outcomes-based approaches that entirely do away with reason. Their focus is placed on the – however wrongful – legally automatic tilting of the takedown scale to favour rights on one side or another, rather than on the reasoning processes through which the scale tilts.
Outcomes-Based Liability Regimes
Automatically favouring freedom of expression, we find the dynamics of section 230 of the US Communications Decency Act, which pioneered a regime of full-fledged exemption of liability for Internet intermediaries, recently replicated in the Brazilian Internet Bill of Rights and, in a more particular context, in the UK Defamation Act 2013.
Such regimes shield intermediaries from liability for keeping online even content they know amounts to a violation of fundamental rights. In effect, knowing or reasoning is immaterial here, for, in the guise of protecting freedom of expression, the law eliminates any encouragement for reasoning by intermediaries at all.
Yet, that may be a somewhat distorted way of protecting freedom of expression. For do these regimes prevent Internet intermediaries from taking content down? Of course they don’t. What these regimes create is, on the contrary, a realm for entirely unrestrained decision making, leaving intermediaries free to reach decisions of whatever quality and purpose they see fit.
There is no legal incentive for best practices and measures of compliance to develop, nor for the assembling of a coherent body of past decisions that we can learn from and make intermediaries accountable for, let alone for the development of feedback loop systems through which intermediaries’ “case law” comes to be reflected in the ways they generally conduct their activities.
Ultimately, here there is just the relinquishment of our rights to absolute sovereigns, who, beyond the forces of markets they anyhow dominate, are bound by no duty to improve our lives in the information environment.
We can call this the Leviathan model.
The existing alternative to regimes of exclusion of liability, however, isn’t any less blunt. What we find in frameworks such as the EU E-Commerce Directive, or in the general law of defamation of different jurisdictions, is a regime with defences that intermediaries can only benefit from if, once acquiring actual knowledge that they host illegal content, they act expeditiously to take such content down.
What those regimes encourage, in other words, is simply the unreflective, automatic takedown of content (and thus the potential violation of freedom of expression) regardless of whether rights would in fact be infringed by the content remaining online. For, indeed, if the content is ultimately found to be illegal, and intermediaries had decided to keep it online, intermediaries won’t be able to benefit from the defences.
Most importantly: it doesn’t matter here how difficult it may have been for intermediaries to interpret the underlying situation – nor how diligent intermediaries might have been in doing so. If they are wrong, they might respond. The only kind of effort that counts here is the effort towards taking the content down. With the sword of Damocles above their heads, it is unsurprising that intermediaries would often simply choose to do so.
Efforts-Based Liability Regimes: A Proposal
An alternative to both extremes above would be a regime that focused not on the default outcomes of tilting the scale in one direction or another, but on the reasoning processes undertaken by intermediaries in making a decision about content they host, on the efforts made by intermediaries in seeking to reach a reasonable interpretation about the factual and normative circumstances entailed by their decisions.
But wouldn’t a regime like this leave intermediaries in an even more uncertain situation? I trust it wouldn’t. On one hand, what would matter here isn’t whether intermediaries are ultimately right or wrong, but whether intermediaries have acted responsibly in seeking to reach the outcomes they have reached – like journalists, who are entitled to make mistakes, as long as they seek responsibly to avoid these.
On the other hand, an efforts-based regime could enable different forms of calibration for intermediary liability. First, such a regime could take into account how intermediaries’ different economic and technological possibilities affect their reasoning processes. As such, the regime could be particularly understanding of new actors’ learning curves, preserving the innovation-inducing spirit of original ‘Good Samaritan‘ defences in this area.
Second, instead of relying on individual damages, an efforts-based regime could rely on more general forms of sanctioning, ranging from fines to warnings and incentives, opening an entirely different – and much more plausible – realm of negotiation between Internet intermediaries and policy authorities.
There are different theoretical rationales for championing an efforts-based regime, which I have explored in recent work. All of them, ultimately, concern a requirement of justice that demands we understand intermediaries as equal members of the community of rights and duties we all inhabit – that there are sufferings which are correlated to their wrongs, and normative consequences of their invisible jurisprudence that transcend the individual cases they settle. It is to this requirement of justice that our expectations towards intermediaries should be tailored. Neither neutrality, nor perfection – what we should expect from intermediaries is, simply, responsibility.
This blog gives the views of the author and does not represent the position of the LSE Media Policy Project blog, nor of the London School of Economics and Political Science. The post introduces an argument explored in more detail by the author in Beyond Gatekeeping: The Normative Responsibility of Internet Intermediaries, 18:4 Vanderbilt Journal of Entertainment & Technology Law 783 (2016).